Agreement between radiological and computer navigation measurement of lower limb alignment

Accurate and reproducible measurements of limb alignment are necessary for planning, performing and evaluation of reconstructive knee surgery. Aim of this study was the comparison of the alignment measured on long-leg standing radiographs with the intraoperative data from a navigation system. The records of 135 consecutive patients who received computer-assisted TKA were examined. Technical quality of the long-leg radiographs (LLRs) was classified good, acceptable or poor according to the rotation of the leg. The difference between radiographic and navigation measurements of leg alignment was assessed. Preoperative LLRs were rated as good 56.3 % (71.1 % postoperatively), acceptable in 37.0 % (20.0 % postoperatively) and poor in 6.7 % (8.9 % postoperatively). The median difference between radiographic and navigation measurements increased with reduced quality of the LLR [good 1.5A degrees (range 0.0A degrees-9.9A degrees), acceptable 2.5A degrees (range 0.0A degrees-15.0A degrees), poor 4.5A degrees (range 0.2A degrees-9.5A degrees)], but not with greater deformity. Median difference between both measurements in good radiographs was 1.7A degrees (range 0.0A degrees-9.9A degrees) preoperatively and 1.2A degrees (range 0.0A degrees-7.0A degrees) postoperatively. Difference between radiographic and navigation measurements of lower limb alignment is low if the LLR are obtained in neutral rotation. Larger differences between both measurements can occur even under these ideal conditions, and it is still unclear which measurement is closer to reality. Therefore, even if a navigation system is used during surgery, long-leg standing radiographs should currently not be abandoned. III.